Guy HowardProgramme Manager, Water Engineering and Development Centre,Loughborough University, UK
Jamie BartramCo-ordinator, Water, Sanitation and Health Programme, World HealthOrganization, Geneva, Switzerland
Executive summary
The quantity of water delivered and used for households is an important aspect ofdomestic water supplies, which influences hygiene and therefore public health. To date,WHO has not provided guidance on the quantity of domestic water that is required topromote good health. This paper reviews the requirements for water for health-relatedpurposes to derive a figure of an acceptable minimum to meet the needs for consumption(hydration and food preparation) and basic hygiene.
Based on estimates of requirements of lactating women who engage in moderate physicalactivity in above-average temperatures, a minimum of 7.5 litres per capita per day willmeet the requirements of most people under most conditions. This water needs to be ofa quality that represents a tolerable level of risk. This volume does not account for healthand well-being-related demands outside normal domestic use such as water use in healthcare facilities, food production, economic activity or amenity use.
The basic need for water includes water used for personal hygiene, but defining aminimum has limited significance as the volume of water used by households dependson accessibility as determined primarily by distance and time, but also includingreliability and potentially cost. Accessibility can be categorised in terms of service level.A summary of the degree to which different levels of service will meet requirements tosustain good health and interventions required to ensure health gains are maximised isshown in table S1 below.
Table S1: Summary of requirement for water service level to promote healthService level Access measure
Needs metLevel of healthconcern
No access (quantitycollected oftenbelow 5 l/c/d)
More than 1000m or30 minutes totalcollection time
Consumption – cannot be assuredHygiene – not possible (unlesspractised at source)
Very high
Basic access(average quantityunlikely to exceed20 l/c/d)
Between 100 and1000m or 5 to 30minutes totalcollection time
Consumption – should be assuredHygiene – handwashing and basic foodhygiene possible; laundry/bathing difficult to assure unlesscarried out at source
Consumption – all needs metHygiene – all needs should be met
Very low
Table S1 indicates the likely quantity of water that will be collected at different levels ofservice. The estimated quantities of water at each level may reduce where water suppliesare intermittent and the risks of ingress of contaminated water into domestic watersupplies will increase. Where optimal access is achieved, but the supply is intermittent,
a further risk to health may result from the compromised functioning of waterbornesanitation systems.
The public health gains derived from use of increased volumes of water typically occurin two major increments. The first relates to overcoming a lack of basic access, where thedistances and time involved in water collection result in use of volumes inadequate tosupport basic personal hygiene and may be marginally adequate for human consumption.
Further significant health gains occur largely when water is available at household level.Other benefits derived from the second step in improving access include increased timefor example, child-care and food preparation and productive activity. Health gainsderived from increased access between these two major steps appear limited, althoughother gains in relation to increased time for activities such as child-care, food preparationand productive activity (including education) may be significant and progressive. Furtherincremental improvements may also occur at higher levels of service, associated withfurther increased access and drinking-water quality control, but also linked to improvedsocio-economic status.
Where the basic access service level has not been achieved, hygiene cannot be assuredand consumption requirements may be at risk. Therefore providing a basic level of accessis the highest priority for the water and health sectors.
Within the population served by basic levels of service, public health gains are primarilyachieved through providing protected water sources, promoting good water handlinghygiene practices and household treatment of water and in other key hygiene behaviours(notably hand and face washing) at critical times.
The categories of service level can also been understood in terms of household watersecurity, although a full description of this would also require estimates of quality andsafety. The group with no access have no household water security. The group with basicaccess could be described as having partial household water security, with the remaininggroups described as having sustained household water security, dependent on the qualityof water supplied.
The service level categories shown in table 1 should be compared with data concerningestimates of present level of coverage by service level as summarised in table S2 (WHOand UNICEF, 2000)1. These figures show that there remains a significant proportion ofthe world’s population (18%) without access to an improved water supply within onekilometre of their dwelling and that 53% do not have access to an intermediate level ofservice as defined in table S1.
The figures for access to an intermediate level of service of water are lower than forsanitation (60%), for which definitions for reasonable access were all related tohousehold or near-household levels of service. Currently there is, justifiably, significantadvocacy to reduce the sanitation access deficit, however, this evidence suggests that to
1 Improved water supplies were: household connection, public standpipe, borehole, protected dug well,protected spring and rainwater collection.. Unimproved water supplies were: unprotected well,unprotected spring, vendor-provided water, bottled water, tanker truck provision
meet a more health-centred definition of access to improved water supply, equal attentionis required for improvement in both water supply and sanitation.
Table S2: Water supply access data for 1990 and 2000 by no access, access toimproved sources and piped supply (from WHO and UNICEF, 2001)Year No access
(millions)Access to improved sourceswithin 1 kilometer (millions)
Access through householdconnections (millions)
1990 22%(1169)
78%(4086)
43%(2255)
2000 18%(1069)
82%(4988)
52%(3169)
The right to water exists at the level of the individual and implies access to the minimumnecessary for basic needs. Progress towards universal achievement of this level of serviceis associated with substantial health gain and remains a focus on international policyinitiatives through the Millennium Declaration Goals and of monitoring activities throughthe WHO/UNICEF Joint Monitoring Programme.
Where achievement of full access to a basic level of service has not been achieved, policyinitiatives should address increasing the numbers of households with this level of service. Maximum health benefits are likely to be obtained by directing resources towardsensuring that all households have access to improved water sources, and in somecircumstances in directly upgrading to access at household level (generally through pipedmeans). Significant gains are also likely to be achieved by upgrading of those with accessto improved sources to household level access. In contrast increasing ease of access toimproved sources outside the household is likely to provide limited health returns.Assessment of progress towards this level of access should be a target of policy in allcountries and in particular where basic needs have been met. Health and other benefitsfrom improved water supply are significantly greater when there is a supply ofcontinuous access to safe drinking water within the home, a level of service that can bedefined as optimal.
In practice, the use of water for domestic purposes cannot easily be distinguished fromproductive use at the household level, particularly among poor urban communities.Domestic water use to sustain livelihoods among the poor forms an integral part ofhousehold coping strategies. There may also be important health and social gains fromensuring adequate quality of service to support small-scale productive use, for examplewhere this involves food production. Access to water adequate for small-scale productiveactivity in such areas is therefore important as part of poverty alleviation and may deliversignificant indirect health benefits as a result.
cavillona
Inserted Text
1 Introduction.....................................................................................................................12 Defining domestic water supply..................................................................................23 Consumption ...................................................................................................................3
3.1 Basic hydration requirements ..................................................................................33.2 Published Reference Values ....................................................................................43.3 Specific population groups.......................................................................................53.5 Hydration needs: types of fluid intake required....................................................73.6 Quality of water for consumption ...........................................................................83.7 Quantities of water required for cooking ...............................................................8
4 Water quantity requirements for hygiene ................................................................94.1 The links between water supply, hygiene and disease ...................................... 104.2 Relationships between water, sanitation hygiene and diarrhoea..................... 104.3 Relationships between water, hygiene and other infectious diseases.............. 154.4 Minimum quantity of water required for effective hygiene.............................. 164.6 Quantity and accessibility: how much do people use and what are thelinks? ............................................................................................................................. 174.7 Quantity and cost: what influence does this have on use?................................ 194.8 Other factors that may affect quantities of water used...................................... 204.9 Laundry: on and off-plot use ................................................................................ 214.10 Minimum requirements for all hygiene needs .................................................. 22
5 Other Uses of water and links to quantity............................................................. 235.1 Household productive uses of domestic water ....................................................... 235.2Amenity uses of water ............................................................................................... 246 Implications .................................................................................................................. 24
6.1 Changing the nature of the debate: household water security/access notquantity........................................................................................................................... 246.2 International Development Targets...................................................................... 256.3 Global Assessment ................................................................................................. 266.4 Emergencies and disasters .................................................................................... 276.5 Quantity in health-based surveillance programmes........................................... 27
1 IntroductionDomestic water supplies are one of the fundamental requirements for human life. Withoutwater, life cannot be sustained beyond a few days and the lack of access to adequate watersupplies leads to the spread of disease. Children bear the greatest health burden associatedwith poor water and sanitation. Diarrhoeal diseases attributed to poor water supply, sanitationand hygiene account for 1.73 million deaths each year and contribute over 54 millionDisability Adjusted Life Years, a total equivalent to 3.7% of the global burden of disease(WHO, 2002). This places diarrhoeal disease due to unsafe water, sanitation and hygiene asthe 6th highest burden of disease on a global scale, a health burden that is largely preventable(WHO, 2002). Other diseases are related to poor water, sanitation and hygiene such astrachoma, schistosomiasis, ascariasis, trichuriasis, hookworm disease, malaria and Japaneseencephalitis and contribute to an additional burden of disease.
As of 2000 it was estimated that one-sixth of humanity (1.1 billion people) lacked access toany form of improved water supply within 1 kilometre of their home (WHO and UNICEF,2000). Lack of access to safe and adequate water supplies contributes to ongoing povertyboth through the economic costs of poor health and in the high proportion of householdexpenditure on water supplies in many poor communities, arising from the need to purchasewater and/or time and energy expended in collection. Access to water services forms a keycomponent in the UNDP Human Poverty Index for developing countries (UNDP, 1999).
The importance of adequate water quantity for human health has been recognised for manyyears and there has been an extensive debate about the relative importance of water quantity,water quality, sanitation and hygiene in protecting and improving health (Cairncross, 1990;Esrey et al., 1985; Esrey et al., 1991). Despite this debate, international guidelines or normsfor minimum water quantities that domestic water supplies should provide remain largelylacking. For instance, whilst the Millennium Declaration Goals include a target to 'halve theproportion of people who are unable to reach or to afford safe drinking water by 2015' (UN,2000) it does not specify in what quantity such water should be supplied. The WHO/UNICEFJoint Monitoring Programme, which produces the Global Assessment of Water Supply andSanitation data, describe reasonable access as being 'the availability of at least 20 litres perperson per day from a source within one kilometre of the users dwelling' (WHO andUNICEF, 2000). However, it should be noted that this definition relates to primarily to accessand should not necessarily be taken as evidence that 20 litres per capita per day is arecommended quantity of water for domestic use.
Norms for quantities of water to be supplied have been proposed for certain specificconditions. For instance the SPHERE project sets out 15 litres of water used per capita perday as being a key indicator in meeting minimum standards for disaster relief (SPHERE,1998). In their guidance manual prepared for the Department for International Development(UK), WELL (1998) suggested that a minimum criterion for water supply should be 20 litresper capita per day, whilst noting the importance of reducing distance and encouraginghousehold connection. A similar figure has been suggested by other researchers (Carter et al.,1997). Gleick (1996) suggested that the international community adopt a figure of 50 litresper capita per day as a basic water requirement for domestic water supply.
2
Many uses of water occur largely at the household (for instance drinking, eating and hand-washing); others may occur away from the home (laundry and in some cases bathing). Thistherefore needs to be borne in mind when ensuring that adequate quantities of domesticsupply are available for these purposes and in interpreting and applying minimum values.
Despite common claims of WHO standards relating to water quantity, WHO has notpreviously published specific guidance on the quantities of water as targets for the healthprotection and promotion. This is in contrast to the concerted effort made, for example, inrelation to establishing international standards and later guidelines for drinking-water quality(WHO, 1985; 1993), wastewater use (Mara and Cairncross, 1989) and recreational waterquality(WHO, in finalisation).
It is important to distinguish quantities of water required for domestic purposes (whichprimarily influence health and productivity), and quantities of water required for otherpurposes (such as agriculture, industry, commerce, transport, energy and recreation). Overall,the requirements for domestic supply typically constitute a very minor component of totalwater withdrawals (Gleick, 1993; 1996).
The purpose of this paper is to review the evidence of the relationships between waterquantity, access and health and to provide a basis for the establishment of minimum quantityand/or access targets for domestic water supplies. It does not address the requirement ofwaters for specific groups (e.g. athletes), specific settings (e.g. hydration needs during airtravel or particular occupational settings) or health impacts related to hydration derived fromalcohol consumption.
The paper draws on an extensive literature review based primarily on the published literature,but in some cases also drawing on ‘grey’ literature where the data was believed to be of goodquality and where this provided clearer information. Key word searches in CambridgeScientific Abstracts (including Aqualine, Water Resource Abstracts and BacteriologyAbstracts) and Medline were employed. In addition, a review was undertaken of availablematerials (papers, books, theses, conference proceedings) at WEDC and WHO resourcecentres. A representative literature was captured through this process, although as notedwithin the text in some areas available literature and evidence is sparse.
2 Defining domestic water supplyIn its Guidelines for Drinking-Water Quality, WHO defines domestic water as being 'waterused for all usual domestic purposes including consumption, bathing and food preparation'(WHO, 1993; 2002). This implies that the requirements with regard to the adequacy of waterapply across all these uses and not solely in relation to consumption of water. The Guidelinesexclude some specific uses (for instance dialysis and contact lens cleaning) and elevatedrequirements for some particularly sensitive sub-populations (for instance the severelyimmuno-compromised). Although this broad definition provides an overall framework fordomestic water usage in the context of quality requirements, it is less useful whenconsidering quantities required for domestic supply.
Sub-dividing uses of domestic water is useful in understanding minimum quantities ofdomestic water required and to inform management options. In the 'Drawers of Water' studyon water use patterns in East Africa, White et al. (1972) suggested that three types of usecould be defined in relation to normal domestic supply:
3
• Consumption (drinking and cooking)• Hygiene (including basic needs for personal and domestic cleanliness)• Amenity use (for instance car washing, lawn watering).
In updating the Drawers of Water study, Thompson et al. (2001) suggest a fourth categorycan be included of 'productive use' which was of particular relevance to poor households indeveloping countries. Productive use of water includes uses such as brewing, animalwatering, construction and small-scale horticulture.
The first two categories identified by White et al. (1972): ‘consumption’ and ‘hygiene’, havedirect consequences for health both in relation to physiological needs and in the control ofdiverse infectious and non-infectious water-related disease. The third category: ‘amenity’may not directly affect health in many circumstances. Productive water may be criticalamong the urban poor in sustaining livelihoods and avoiding poverty and therefore hasconsiderable indirect influence on human health (Fass, 1993; Thompson et al., 2001).
The different primary uses of water are discussed in the following sections and the quantityrequirement of each and its implication for health are reviewed.
3 ConsumptionWater is a basic nutrient of the human body and is critical to human life. It supports thedigestion of food, adsorption, transportation and use of nutrients and the elimination of toxinsand wastes from the body (Kleiner, 1999). Water is also essential for the preparation offoodstuffs and requirements for food preparation are included in the discussion ofconsumption requirements.
3.1 Basic hydration requirementsThe human body requires a minimum intake of water in order to be able to sustain life beforemild and then severe dehydration occurs. Adverse health effects have been noted from bothmild and severe dehydration and the latter can be fatal.
The US National Institutes of Health (2002) provide a definition of mild dehydration as beinga loss of 3-5% of body weight, moderate dehydration as being 6-10% loss of body weight andsevere dehydration (classed as a medical emergency) 9-15% loss of body weight. In a recentreview Kleiner (1999) defined mild dehydration as being the equivalent of 1-2% loss of bodyweight through fluid losses and over 2% loss as severe dehydration, whilst noting that there isno universally applied index of hydration status. Mild dehydration can be reversed byincreased fluid intake and this may be enhanced through the use of salt replacement solutions.Severe dehydration will require rehydration strategies involving more than simple fluidreplacement, and often food or other osmolar intake is needed; the process may take up to24 hours (Kleiner, 1999).
Dehydration may be a short-term effect, for instance resulting from loss of body fluids insevere diarrhoea, which can be fatal. Short-term dehydration may also result from excessalcohol intake or increased water loss due to increased temperature and altitude or decreasedrelative humidity combined with inadequate fluid replacement. Dehydration may also belong-term (often mild) which may result in adverse health effects (Chan et al., 2002; Kleiner,
4
1999). Long-term dehydration may result from inadequate fluid replacement, often as aconsequence of depressed thirst mechanisms and perceptions of poor beverage taste.
Mild dehydration has been associated with a number of adverse health effects, includingincreased risks in susceptible groups to urinary stone formation, increased risks of urinarytract cancer and poor oral health. Urinary stone formation is significantly increased when theurine volume excreted is below 1 litre per day; urinary volumes exceeding 2 to 2.5 litres perday can prevent recurrence of stones in previously affected patients (Kleiner, 1999). White etal (1972) suggest that in order to reduce the risk of kidney stones, a minimum of 1.5 litresshould be passed as urine each day.
A recent study in the Adventist community in California noted a strong negative associationwith intake of water and the risk of fatal coronary heart disease for both men and women(Chan et al., 2002). Relative risks for men reduced to 0.46 for high volume water intake (5 ormore glasses) and 0.54 for women who had medium intake (3-4 glasses) compared to lowwater intake (2 or less glasses). If it is assumed that each glass contained 0.25 litres (areasonable estimate of size of glass) an estimate of a minimum of 1.25 litres per capita perday for men and 0.75 litres per capita per day for women is required to reduce risks of fatalcoronary heart disease. Taking an average of these figures provides 1.0 litre per capita perday for a population-based estimate of the volume of water that reduces the risk of fatalcoronary disease. Some studies have also indicated decreased risks of colonic and breastcancer with increasing fluid intake (Kleiner, 1999).
Kleiner (1999) notes that in general there is less available information regarding adverseeffects on cognitive performance by dehydration, but highlights three studies suggesting thatthis would occur. A further review of literature indicated limited available studies in this areaand somewhat contradictory evidence. In a study using volunteers, Neave et al., (2001) foundno significant main effects of hydration status on cognitive performance, whilst notinggreater mood alertness after successive drinks. By contrast, Rogers et al. (2001) suggestedthat there was an immediate (although not sustained) improvement in cognitive performanceon ingestion of water. This suggests that this area requires further investigation.
It is pertinent to note that the majority of health effects derived from dehydration are derivedprimarily from developed countries and there is very little available data from developingcountries. However, although these impacts may represent an overall lower proportion of theburden is disease in developing countries, the effects from dehydration would not be expectedto be different in developing counties.
3.2 Published Reference ValuesIn their review, White et al. (1972) suggested that 2.6 litres of water per day is lost throughrespiratory loss, insensible perspiration, urination and defecation. In addition, a significantquantity of water is lost through sensible perspiration if hard work is performed. Thesefigures led them to suggest that a daily minimum of water required in tropical climates wouldbe around 3 litres per person, although the volume of water loss suggests that this should beat the upper end of this scale. They note, however, that under extreme conditions of hardwork at high temperatures in the sun this figure could rise to as much as 25 litres per day.However, they also point out that the proportion of the fluid intake achieved via food wouldbe expected to vary significantly and could provide 100% of the fluid requirement in somerare cases, notably pastoralists where milk was the primary food.
5
Kleiner (1999) suggests that, based on US National Research Council guidelines in relation tohydration needs resulting from average energy expenditure and environmental exposure inthe USA, the average male should consume a minimum 2.9 litres per day and the averagefemale 2.2 litres. Approximately one-third of this fluid was considered likely to be derivedfrom food.
In the WHO Guidelines for Drinking-Water Quality, Guideline Values for chemicalcontaminants are based on the assumption of a 60 kg adult consuming 2 litres per day fromdrinking water, which would be equivalent to 3 litres per capita per day including foodconsumption (if the ratio cited by Kleiner were applied). Where specific guidance is neededfor vulnerable populations, a figure of 1 litre per day for a 10kg child or 0.75 litre per day fora 5kg child are used (WHO, 1993; p31). The WHO-UNEP-ILO International Programme onChemical Safety use reference values for volume of fluid intake in deriving its guidance,using reference body weights of 70kg for adult males, 58kg for adult females and an averageof 64kg. The reference fluid intake values for these different reference body weights underdifferent climatic and activity conditions are shown in table 1 below.
Table 1: Daily fluid intake reference values in litres per capita (IPCS, 1994)
3.3 Specific population groupsParticular population groups have particular hydration requirements, including youngchildren, pregnant or lactating women, the elderly, the terminally ill and athletes. Athletes arenot discussed in this review as hydration would typically also include salt replacement andboth the cause of dehydration and management of rehydration are more specialised than maylegitimately be expected from a normal domestic supply.
As rehydration primarily relates to the replacement of lost fluid from natural processes, it isimportant to consider the losses of fluid from different age groups when consideringvulnerable sub-populations. The losses of water from the bodies of small children areproportionally considerably greater than for adults, 15% of fluid per day as opposed to 4%.These proportionately higher losses explain why a 7 kg child requires 1 litre per day fluid toreplace lost fluid compared to 2.9 litres for a 70kg adult male, the increase in replacementfluid being a factor of three compared to a 10-fold difference in weight (Kleiner, 1999). Low-birth-weight infants need proportionally even greater fluid replacement per kilogram ofweight than do other infants (Roy and Sinclair, 1975). The figures for fluid replacementexceed those for fluid intake noted above because this includes all forms of fluid used inhydration, approximately 30% of which will be derived from food rather than consumption ofliquid.
Pregnant women also require additional fluid replacement to ensure that foetal needs are met,as well as providing for expanding extra-cellular space and amniotic fluid. The US National
6
Research Council suggests an allowance of an extra 30ml per day during pregnancy (Foodand Nutrition Board, 1989). Lactating women have additional water requirements, leading toan additional requirement of 750ml to 1 litre per day for the first six months of lactation(Food and Nutrition Board, 1989). In those groups with least access to water supply, womenwho are pregnant or lactating frequently continue to undertake at least moderate activity inhigh temperatures, therefore these needs are additive to the basic needs of all adults.
The elderly may not require additional volumes of water, but may be at greater risk fromdehydration due to decreasing thirst sensations (Phillips et al., 1984). Furthermore, studieshave noted a relationship between age and the ability of the body to concentrate urine,suggesting an increasing water requirement to maintain good renal functioning (Rowe et al.,1976).
For the terminally ill, Jackonen (1997) highlights a range of benefits and burdens related todehydration, with benefits accrued from lower levels of distress and lower awareness of painand reduced requirements for urination with the pain and discomfort that this may cause.Benefits of hydration include preventing dehydration and malnourishment as well asprolonging life and avoiding health problems such as renal failure. The benefits and burdensassociated with dehydration amongst the terminally ill often relate to medical hydration(intravenous, nasogastric or nutrition administration) and therefore will have little impact onvolumes of water required in a general domestic supply.
3.4 Requirements to maintain hydrationThe evidence outlined above indicates that there are health concerns regarding hydrationstatus and therefore the requirements for human to maintain an adequate hydration level andminimise risk of disease associated with adverse effects outlined above must be defined. Thedefinition of the ‘absolute minimum’ quantity of water to sustain hydration remains elusive,as this is dependent on climate, activity level and diet.
In developing countries, White et al. (1972) and Gleick (1996) suggest that a minimum of 3litres per capita per day is required for adults in most situations. However, households withleast access to water supplies are more likely to be engaged in at least moderate activity andoften in above-average temperatures. Data from the US Army reported in White et al. (1972)provides estimates of water quantity needs at different temperatures and activity levels. Thisindicates that at 25oC with moderate activity in the sun (for instance agricultural work)approximately 4.5 litres are required to maintain hydration. This rises to about 6 litres at 30oCor when hard work in the sun is undertaken at 25oC. Although the US Army has more recentrecommendations for hourly intake of water per hour in relation to heat categories andactivity intensity to prevent heat injury, this do not easily translate into non-military activity.They do, however, stipulate that hourly fluid intake should not exceed 1.08 quarts (1.03litres) and that daily intake should not exceed 12 quarts (11.35 litres) (United States ArmyCenter for Health Promotion and Preventive Medicine, 2003).
The literature reviewed in section 3.2 to 3.4, indicates that the quantity of water required forhydration (whether via direct ingestion or food) should be a minimum of 2 litres for averageadults in average conditions, rising to 4.5 litres per day under conditions typically facing themost vulnerable in tropical climates (see table 2 below) and higher in conditions of raisedtemperature and/or excessive physical activity. This figure can be interpreted as applying toall adults and to children, given the difficulty in determining whether the ration of adult/childwater requirements would remain the same with increasing activity and/or temperature. These
7
values encompass the range in which beneficial impacts on prevention of coronary diseaseand kidney stone occurrence appears likely and would be at the lower end of requirements toprevent recurrence of kidney stones.
By also taking into consideration the needs of lactating women, many of whom in the groupwith least access will still be expected to undertake moderate activity in high ambienttemperatures, a minimum quantity required of fluid required for hydration (via both directconsumption and food) can be estimated as 5.5 litres per capita per day. This will not accountfor those in unusually hot environments or engaged in strenuous physical activity whereminimum needs may be considerably greater.
As some hydration needs are met through fluid obtained from food, the figure of 5.5 could beinterpreted in two ways. Firstly, it could be assumed that the water supply should be able tomeet all hydration needs (i.e. minus contributions from food). The second approach would beto assume that one-third of all hydration fluid is derived from food and that thereforedomestic water supply need only meet two-thirds of the minimum quantity identified.
In this report we use the former approach because as the proportion of fluid obtained fromfood may vary significantly in response to diet and culture from negligible to all hydrationneeds as noted by White et al. (1972). Therefore, trying to allocate a proportion of the fluidrequirement to food on a global basis would risk significantly under-estimating waterquantity needs in a number of situations where food contribution is negligible. As these arelikely to be found in situations with vulnerable populations, notably in emergencies, thiswould entail a serious risk. Allocating the full hydration component to drinking water mayover-estimate the quantity of water required, but this is believed to be no more significant thatthe variation likely to occur due to activity levels and temperature.
Table 2: Volumes of water required for hydration
Volumes (litres/day)Average
conditionsManual labour inhigh temperatures
Total needs inpregnancy/lactation
Femaleadults
2.2 4.5 4.8 (pregnancy)5.5 (lactation)
Male adults 2.9 4.5 -
Children l.0 4.5 -
3.5 Hydration needs: types of fluid intake requiredThe benefits derived from specific types of fluid consumed are a matter of some debate. Forinstance, Kleiner (1999) suggests that drinking diuretics such as coffee may lead to milddehydration, although a preliminary study by Grandjean et al. (1999) suggests that there is nosignificant difference between use of different beverages on hydration status, whilstindicating that further research is required.
Chan et al. (2002) demonstrated a statistically significant positive association betweenconsumption of fluids other than water and increased risk of coronary disease among womenconsuming over 5 glasses per day compared to those consuming less than 2 glasses per day.The relative risk was 2.47, although the confidence interval were wide. However, the pointestimates of relative risk did not change even when adjusting for more traditional factors for
8
coronary disease. A positive association was also noted in men, although this was notstatistically significant. An attempt was made to identify the impact of specific fluids, but thiswas not possible for juices and sugared drinks because of generally low daily intake, theconsumption of caffeinated beverages was positive but not statistically significant and theintake for soy milk was close to null. The authors note that the population studied consumedless alcohol and caffeine than the USA average which may limit the ability to project thefindings more broadly. Although the analysis from this study does not support riskassessments of specific fluids, the study does highlight the specific value of water as opposedto more generic fluid intake.
3.6 Quality of water for consumptionThe quality of water that is consumed is well-recognised as an important transmission routefor infectious diarrhoeal and other diseases (WHO, 1993). The importance of water qualitycontinues to be emphasised by its role in epidemics and contribution to endemic disease frompathogens (Ford, 1999; Payment and Hunter, 2001). This affects both developed anddeveloping countries, although the majority of the health burden is carried by children indeveloping countries (Prüss et al., 2002; WHO, 2000). However, recent outbreaks such asthat of cryptosporidiosis in Milwaukee and E.coli O157:H7 and Campylobacter jejuni inWalkerton, Ontario illustrate that the developed world also remains at risk (Mackenzie et al.,1994; O’Connor, 2002).
Disease may also result from consumption of water containing toxic levels of chemicals. Thehealth burden is most significant for two chemicals: arsenic and fluoride. Arseniccontamination of drinking water sources is being found in increasing numbers of watersupplies world-wide and in Asia in particular. The total disease burden is as yet unknown, butin Bangladesh, the country with the most widely reported problem, between 35 and77 million people are at potential risk (Smith et al., 2000). Fluoride is also a significantglobal problem and WHO (1999) suggest that over 60 million people are affected by fluorosisin India and China and suggest the total global population affected as being 70 million.Nitrate is also of concern although there remains uncertainty about the scale of adverse healtheffects from nitrate as few countries include methemaglobinaemia as a notifiable disease(Saywell, 1999). Raised nitrate is, however, identified as a potential public health problems incountries where concentrations in groundwater reach extremely high values (Melian et al.,1997).
Water provided for direct consumption and ingestion via food should be of a quality that doesnot represent a significant risk to human health. A 'zero-risk' scenario for public supplies isnot achievable and evidence points to the need to define tolerable risks, commonly based onestimates of numbers of excess cases per defined population size. This approach underpinsmuch risk assessment thinking within the water sector for both microbial and chemicalcontaminants (Fewtrell and Bartram, 2001; Haas et al., 1999; WHO, 1996).
3.7 Quantities of water required for cookingWater is essential as a medium for preparing food. One study noted that the volume ofcooking water available may be an important determinant for diarrhoea incidence in childrenover 3 years of age, although this was less important than water quality for the under 3 yearsage group (Herbert, 1985).
9
Defining the requirements for water for cooking is difficult, as this depends on the diet andthe role of water in food preparation. However, most cultures have a staple foodstuff, whichis usually some form of carbohydrate-rich vegetable or cereal. A minimum requirement forwater supplies would therefore also include sufficient water to be able to prepare an adequatequantity of the staple food for the average family to provide nutritional benefit.
It is difficult to be precise about volumes required to prepare staples as this depends on thestaple itself. However, an example can be provided for rice, which probably represents themost widely used staple food worldwide. Recommendations for nutrition usually deal withthe intake of nutrients rather than specific food stuffs. Most food pyramids give a suggest anintake for cereals of 6 to 11 servings per day, or 600 – 1100 grams per day (GraemeClugston, personal communication). To prepare rice using the adsorption method (i.e. onlysufficient water to cook the rice is added), 1.6 litres is required for 600g per capita per day.
More water may be required to ensure that other foodstuffs can be cooked, although definingminimum quantities is difficult as this depends on the nature of the food being prepared. Forinstance, Gleick (1996) suggests that on average 10 litres per capita per day is required forfood preparation, whilst Thompson et al. (2001) show that in East Africa only 4.2 litres percapita per day were used for both drinking and cooking for households with a pipedconnection and even less (3.8 litres per capita per day) for households without a connection.Taking into account drinking needs, this suggests that between 1.5 and 2 litres per capita perday is used for cooking.
If the quantity of water required for cooking rice is taken as representing the needs for staplepreparation and assuming further water is required for preparation of other food, the evidencesuggests that in most cases approximately 2 litres per capita per day should be available fromdomestic supplies to support food preparation.
By adding the volume required for food preparation to the volumes identified in table 2, afigure for total consumption (i.e. drinking water plus water for foodstuffs preparation) of7.5 litres per capita per day can be calculated as the basic minimum of water required, takinginto account the needs of lactating women.
4 Water quantity requirements for hygieneThe need for domestic water supplies for basic health protection exceeds the minimumrequired for consumption (drinking and cooking). Additional volumes are required formaintaining food and personal hygiene through hand and food washing, bathing and laundry.Poor hygiene may in part be caused by a lack of sufficient quantity of domestic water supply(Cairncross and Feachem, 1993). The diseases linked to poor hygiene include diarrhoeal andother diseases transmitted through the faecal-oral route; skin and eye diseases, in particulartrachoma and diseases related to infestations, for instance louse and tick-borne typhus(Bradley, 1977; Cairncross and Feachem, 1993).
The relative influence of consumption of contaminated water, poor hygiene and lack ofsanitation on diarrhoeal disease in particular has been the topic of significant discussion (seefor example Esrey et al., 1985). This has mirrored a broader debate within the health sectorworldwide regarding the need for quantifiable evidence in reducing health burdens. Thedesire for evidence-based health interventions is driven by the need to maximise benefitsfrom limited resources (a critical factor both for governments and their populations). It is also
10
driven by the desire to ensure that populations benefit from the interventions that deliver thegreatest improvement in their health.
4.1 The links between water supply, hygiene and diseaseClassifying diseases by causative agent such as microbe type for infectious disease has avalue in terms of understanding aetiology of infection. However, a more effective way toinform decision-making is to categorise pathogens /diseases in relation to the broad mode oftransmission. Bradley (1977) suggests that there are four principal categories that relate towater and which are not mutually exclusive:
• water-borne - caused through consumption of contaminated water (for instance diarrhoealdiseases, infectious hepatitis, typhoid, guinea worm);
• water-washed - caused through the use of inadequate volumes for personal hygiene (forinstance diarrhoeal disease, infectious hepatitis, typhoid, trachoma, skin and eyeinfections);
• water-based - where an intermediate aquatic host is required (for instance guinea worm,schistosomiasis); and,
• water-related vector - spread through insect vectors associated with water (for instancemalaria, dengue fever).
Other workers have suggested a change in this classification system to replace the waterbornecategory with faecal-oral (to reflect multiple routes of transmission) and to restrict the water-washed diseases to only as those skin and eye infections that solely relate to the quantity ofwater used for hygiene (Cairncross and Feachem, 1993). The original Bradley (1977) systemhas particular value as its focus is on the potential impact of different interventions. Theoccurrence of particular diseases in more than one group is a legitimate outcome wheredistinct interventions may contribute to control. Thus guinea worm for example is classifiedas both a water-based disease and water-borne disease.
4.2 Relationships between water, sanitation hygiene and diarrhoeaDiseases primarily transmitted through the faecal-oral route (shown in figure 1 below)include infectious diarrhoea, typhoid, cholera and infectious hepatitis. Faecal-oral diseasesare associated with acute symptoms (with a probability of death) and in some cases withdelayed sequelae. Transmission may occur through a variety of mechanisms, includingconsumption of contaminated water and food as well as through person-person contact(Bradley, 1977). These are dealt with together here, in order to emphasise the importance oflocal disease patterns rather than applying generic models. The available evidence fromhealth studies suggests that interventions are likely to be locality-specific and are determinedby timing and the interaction between different factors. As noted by Vanderslice and Briscoe(1995) as all the interventions deliver some improvement, the relative impacts of each mayhave limited relevance for policy.
Other factors apart from water and sanitation facilities and hygiene behaviours maysignificantly influence diarrhoeal disease. For example breast-feeding has been noted inseveral studies as being protective against diarrhoeal disease independently of otherinterventions (Al-Ali et al., 1997; Vanderslice and Briscoe, 1995).
11
Figure 1.1: Faecal-oral route transmission of disease (taken from WELL, 1998)
Whilst numerous studies have been undertaken to investigate the relative importance ofdifferent interventions, relatively few have been sufficiently rigorous to allow quantifiedreductions in disease to be estimated. In a review based on 67 studies, Esrey et al. (1985)investigated the relationships between diarrhoeal disease and a number of factors includingwater quality, water availability and excreta disposal. The findings from this review aresummarised below in table 3 and suggest that median reductions in diarrhoeal disease fromwater availability were higher than those recorded for water quality improvements. Combinedimprovements in quality and availability led to greater median reductions in diseaseincidence. The variation in reduction from each intervention type was significant, with thehighest recorded reductions ranging from 48% to 100%. For all interventions, studies werereported that led to no measurable reduction in diarrhoeal disease.
12
Table 3: Percentage reductions in diarrhoeal disease rates attributed to water or excretadisposal improvements (Esrey et al., 1985)
Percentage reductionType of intervention Number of studies Median RangeAll interventions 63 22 0-100Improvements in water quality 9 16 0-90Improvements in water availability 17 25 0-100Improvements in water quality &availability
8 37 0-82
Improvements in excreta disposal 10 22 0-48
In a subsequent review of 144 studies looking at various different single and multipleinterventions in water and sanitation by Esrey et al. (1991), relatively few (56) were deemedto be rigorous and only 24 could be used to calculate morbidity reductions. The findings fromthis study are shown in table 4. The data from the rigorous studies suggested that medianreductions in morbidity were relatively low from all water improvements, unless these werecombined with sanitation improvement. In this review, the impact of combinedimprovements in water quality and quantity resulted in a lower reductions than for waterquantity interventions alone, which is counter-intuitive and contradicts the findings of theprevious review. The authors of this study also noted that the benefits derived from increasedwater availability were not necessarily felt in all age groups, a finding highlighted in aprevious study by Herbert (1985).
Table 4: Reduction in diarrhoeal disease morbidity in one or more components ofwater and sanitation (Esrey et al., 1991)
These findings are most readily understood as demonstrating the significant impact of allinterventions and that the scale and relative impact of a single intervention depends stronglyon the dominant route of exposure under local circumstances. Prüss and Havelaar (2001) notethat for many infectious diarrhoeal diseases, exposure-risk relationships are poorlyunderstood and therefore there are profound difficulties in attributing outcomes to specificexposures. Sazawal et al.(1991) demonstrated that children with persistent diarrhoea have anoverall higher number of episodes of diarrhoeal disease, compared to those suffering fromacute diarrhoea. Persistent diarrhoea will affect immune competency and increase subsequentsusceptibility. Therefore, in understanding local disease patterns, identifying whether asignificant proportion of children suffer from persistent diarrhoea may be important inunderstanding exposure routes.
13
A variety of findings have been reported in other studies from developing countries lookingat diarrhoeal disease incidence. Esrey (1996) concluded on the basis of a review of data fromDemographic Health Surveys from 11 countries that improvements in sanitation led to moresignificant reductions in diarrhoeal disease than improvements in water supply. In thePhilippines, increasing risks of hospitalisation due to severe diarrhoea was noted withdeclining hygiene standards, but only indices of overall cleanliness (a composite measure ofenvironmental hygiene and personal appearance) and kitchen hygiene were significantlyassociated with diarrhoea (Baltazar et al., 1993). Bukenya and Nwokolo (1991) showed thatthe presence of a standpipe at households in urban Papua New Guinea was associated withless diarrhoea than users of communal sources and that this was found across all socio-economic groups, whilst noting the importance of removal of both human and pig faeces.Gorter et al. (1991) showed that in Nicaragua, that children living in households with a watersupply within 500m had 34% less diarrhoea than children living in households whose watersource was over 500m from the house. Gorter et al. (1991) note there once a water supply iswithin 500m, reducing distance had no further impact.
Other studies have assessed the impact of different water supply and sanitation interventionsusing different outcome measures. A study in India that used nutritional status as a healthmeasure suggested that water quality was the principal determinant for health in childrenunder the age of 3, whereas water quantity was most important for children above 3 (Herbert,1985). A previous study in rural Lesotho, height for weight scores were found to be betterassociated with latrine use than water quantity, water quality not being analysed as it hadpreviously been found to be insignificant (Esrey et al., 1992). However, in South India,defecation practice was not a significant predictor of height for weight scores (Herbert,1985).
A recent study in Pakistan demonstrated that increased quantities of water available athouseholds was critical in preventing stunting (van der Hoek et al., 2002). This study showedthat the presence of storage containers linked to household connections was the mostprotective level of service and that limited benefits were accrued when water was onlyavailable via a yard tap compared to collection from a communal source.
The evidence from the literature suggests that water availability has an important influence ofhealth and diarrhoea incidence in particular, although as noted by both Esrey et al. (1991) andHerbert (1985) this is not necessarily true for all age groups. Esrey (1996) suggests that it isonly when the water supply is delivered on-plot that health gains are found. It is also notedthat this may be due to a number of factors, not least of which is the better socio-economicstatus of households with this level of service and possibly better quality of water supplied(Esrey, 1996). There appears to be limited published literature on the impact of providingnon-piped water supplies on-plot, which would be of particular importance in determining thefraction of diarrhoeal disease that is directly attributable to increased service level and thefraction attributable to other factors. In the study reported by Gorter et al. (1991), althoughnot explicitly stated, there is an implication that all water sources when within 500m offeredhealth benefits as no difference was noted in diarrhoeal disease incidence between watersources.
The impact of water availability may have particular benefits for child health. Prost andNégrel (1989) suggest that the quantity of water used for children’s hygiene is sensitive toavailability and that reducing the time taken to collect water (including journey and waitingtime) from 5 hours to 15 minutes, results in 30 times more water being used for child
14
hygiene. It is suggested that reducing the time take to collect water will also allow greatertime to be available for child feeding, food preparation or more frequent feeding as well asbetter hygiene. Other commentaries suggest that reductions in diarrhoeal disease withincreased service level may in fact be much more modest (Vanderslice and Briscoe, 1995).
Despite the evidence pointing to the benefits of increased quantities of water on health, therelationship is not simple and most research has made significant assumptions about wateruse. Hygiene is not solely related to availability of water, but also to specific hygienebehaviours such as hand washing at critical times, for instance before eating and cooking andafter defecation (Cairncross, 1993; Petersen et al., 1998; Shahid et al., 1996; Sircar et al.,1987; Stanton and Clemens, 1987). Factors other than handwashing hygiene may influencethe contamination of hands, of which humidity and temperature are reported to be important(Pinfold, 1990).
In relation to hygiene practices, a review by Huttly et al. (1997) suggested that targetingprogrammes on a single behaviour was more effective than addressing several behaviourswhen promoting hygiene. They suggest that a median reduction of 35% in diarrhoeal diseasemorbidity from improved hand-washing is achievable through well-designed hygieneeducation programmes, with a range of 30-89%. Recent work appears to indicate that themedian reduction may in fact be higher, although the results remain to be published (ValCurtis, personal communication). It should be noted that the figures quoted by Huttly et al.(1997) are of a similar order of magnitude to estimates of disease reduction using householdtreatment of water (Sobsey, 2002). This indicates that household-targeted interventionsdeliver significant improvements in health even when environmental conditions and servicesare not conducive to improved health, although as with all environmental interventions therange of impacts is considerable. In the case of household water treatment, this appears to befrom zero reduction in diarrhoea incidence to about 20% reduction.
The timing of hand washing maybe important. Experience suggests that the most criticaltimes are following defecation and before eating. Curtis et al. (2000) suggest that the criticaltime is post-defecation rather than before eating, while other studies suggest that the reverseis true in some situations (Birmingham et al., 1997). Stanton and Clemens (1987) foundreduction in diarrhoea incidence among young children was influenced by maternal handwashing prior to food preparation.
A number of studies suggest that hand washing with soap is the critical component of thisbehaviour and that hand washing only with water provides little or no benefit (Cairncross,1993; Ghosh et al., 1997; Khan, 1982; Oo et al., 2000). Hoque and Briend (1991) showedthat whilst less effective than when using a rubbing agent, such as soap, mud or ash, somereductions in contamination were found when washing with water alone, but that use ofalternative rubbing agents (mud or ash) provided the same benefits as soap. Hoque et al.(1995) also found that use of mud, ash and soap all achieved the same level of cleanlinesswith hand washing. and suggest that it is the action of rubbing of hands that was moreimportant than the agent used. These authors suggest that rinsing with 2 litres of clean waterwas also protective, although this seems to be a large quantity, which would difficult tosustain in the absence of on-plot access to water.
In a study in Peru, Gilman et al. (1993) found a positive relationship between the quantity ofwater available in the home and the frequency of hand washing in a shantytown. The exactnature of the water supply is unclear, but appears to have been concrete tanks at households
15
that were periodically filled, either from tanker trucks or piped water. The authors of thisstudy concluded that hygiene education was of limited value unless water supplies wereimproved.
4.3 Relationships between water, hygiene and other infectious diseasesInfectious diseases of the skin (a sub-set of water-washed diseases) and trachoma areamongst the diseases on which water quantity would be expected to exert significantinfluence. Trachoma is the most extensively studied disease, given its relatively high impacton health.
In a review of available evidence, Prüss and Mariotti (2000) found that only two studies outthe 19 reviewed showed a significant relationship with water quantity and suggested that therelationship between trachoma incidence and quantity of water was not simple. One study insouthern Morocco that showed a difference in incidence in trachoma between the use of less5 litres per day and use of more than 10 litres per day. Another study was noted as indicatingno relationship to quantity of water used, but which showed a strong relationship betweenavailability of water and trachoma. In another study from Latin America, the protective effectof water quantity was seen only at high levels, as the difference in incidence was related touse of more or less than 5000 litres per capita per month (165 litres per capita per day), (Lunaet al., 1992). Such volumes of water are indicative of water piped into the home. Prüss andMariotti (2000) also note six studies that showed a positive relationship between increasedaccess to water and reduced incidence of trachoma, with a median reduction of 27%, with arange of 11-83% reduction.
In most studies, distance from primary water source to home appears to be the mostsignificant water supply factor influencing trachoma. In a review by Esrey et al. (1991), fourstudies were found that demonstrated a median reduction of 30% in trachoma incidence withshorter distances to the home and two studies that showed no relationship. In all cases, thedifferences in distance associated with significant differences in incidence were againrelatively gross and included:
• household connection versus source more than 500m from household;• water source less than 5 minutes compared to over one hour; and,• water source less than 30 minutes compared to source over 2 hours away.
Only in one study was there a more sensitive measure, the difference being between sourcesmore than or less than 200m distant. All these studies were included within the review byPrüss and Mariotti (2000) and were in the group of six studies showing a reduction intrachoma incidence with increased access.
The reviews by Esrey et al. (1991) and Prüss and Mariotti (2000) provide evidence toindicate that only relatively gross differences in water quantity, as indexed by access byservice level, are significant in relation to the incidence of trachoma. The importance ofdistance is supported by several other studies and reviews on trachoma prevalence (Hsieh etal., 2000; Prost and Négrel, 1989; West et al., 1989). West et al. (1991) concluded that percapita water availability was not associated either with trachoma or facial cleanliness. Baileyet al.. (1991) showed that the reduced volumes of water used for washing children’s faceswere associated with trachoma, but total per capita volume use was not. Several workers havesuggested that it is value placed on the use of water for hygiene rather than source proximity
16
and per capita water collection, which is protective against trachoma (Bailey et al.,1991;West et al., 1989; Zerihun, 1997).
Most studies of the incidence of trachoma suggest that it is hygiene behaviour that is theprimary determinant. For instance, facial cleanliness is noted as being important and appearsto function independently of water quantity (West and al., 1989). Furthermore, clustering ofcases within communities is noted as important, as is the presence of siblings with trachomawithin individual households (West et al., 1991; 1996). Research suggests that factors such ascattle ownership are also important in rural areas and that sanitation and garbage disposal areimportant in both urban and rural areas (Hsieh et al., 2000; West et al., 1996; Zerihun, 1997).This is supported by a review of the evidence to support environmental and facial cleanlinessinterventions in integrated trachoma programmes (Emerson et al., 2000).
Infectious skin diseases have been less well researched. In a review of influences on skindisease in rural Tanzania, Gibbs (1996) found that household density and low socio-economic conditions were important determinants for incidence of transmissible skin disease.The study compared two villages and found that distance to water source was notsignificantly associated with skin disease, despite one village having a water source within 20minutes and the other having a water source that was 46 minutes from the village. Usingfigure 2, this would suggest a significant difference in quantities of water collected.
Evidence from Nepal suggests that the influence of water on hygiene may also reflect theavailability of heated water rather than all water in cold climates. A review of water supplyaccess data and diarrhoeal and skin diseases showed that in Districts which have typicallyvery cold seasons for part of the year, hygiene-related diseases remain much higher thanaverage (Howard and Pond, 2002). It is suggested that despite increasing access to watersupply and sanitation, the absence of hot water inhibits good hygiene, a finding in line withresearch into emergency responses in countries with distinct cold seasons (Buttle and Smith,1999).
4.4 Minimum quantity of water required for effective hygieneThe evidence from the literature consistently points to use of water as being important tocontrolling disease and the fact that lack of access to water may impede its use and therebyadversely affect health. The evidence indicates that the benefit from increased quantity ofwater would only be felt in relation to the gross differences of service level and that hygienebehaviour is more important within populations using communal water sources. This suggeststhat incremental benefits among households with a communal source of water fromincreased quantities of water used are limited.
Review of the data regarding hygiene practices and disease therefore does not enable thedefinition of a minimum quantity of water recommended for use to ensure effective hygieneis practised. The evidence suggests that while benefits are accrued from increased availabilityof water, this is not solely related to quantities of water used, although increased availabilityis likely to increase quantities used. The evidence further suggests that it is the effective useof both water and cleansing agents and the timing of hygiene practices that are moreimportant than volumes of water used. Furthermore, as some hygiene behaviour protective ofhealth, for example laundry and bathing, may be carried out off-plot, the quantity of waterrequired to sustain good hygiene may vary significantly with water collection behaviour.
17
For water quantity to act as an absolute constraint on hygiene, it must be available only invery small quantities. To act as a positive driver for improved hygiene, water must beavailable at higher service levels and ideally supplied at least through one tap on the houseplot. Since (as discussed below), the quantity of water collected may be relatively inflexiblebeyond gross differences in supply type/service level defining a minimum quantity of wateris neither supported by evidence nor of practical value.
4.5 Water quantity/access and non-infectious diseaseThere are other health benefits of increased access within the range implying physicalcollection, notably reduced potential for damage to the spine and for the early onset arthriticdiseases and protection against hip damage (Dufault, 1988; Page, 1996). Where women mustwalk long distances this may exacerbate malnourishment and also affect the quantity andquality of milk produced by lactating women (Dufault, 1988).
4.6 Quantity and accessibility: how much do people use and what are the links?The importance of water availability has been shown above to influence health in thepreceding discussion (sections 3 and 4) and some of the difficulties in interpreting availabledata have been noted. . As noted in section 4.4, the benefit from increased quantity of wateris only felt in relation to the gross differences of service level. It is therefore useful to reviewthe evidence of the interaction between accessibility and quantity in order to assess howhealth may be influenced by these factors.
Most of the studies cited in the above-mentioned reviews as providing evidence of the greaterimportance of quantity actually provided measures of accessibility, with the assumption thatincreased accessibility equates to increased volumes of water used (Esrey et al., 1991). It istherefore important to assess to what extent this relationship is true and what changes inquantity can be expected. This has particular value in ensuring that the formulation of the truenature of the problem is correct in order to inform effective decision making.
Cairncross (1987) provides an example from Mozambique that demonstrated that waterconsumption in a village with a standpipe within 15 minutes was 12.30 litres per capita perday compared 3.24 litres per capita per day in a village where it took over five hours tocollect a bucket of water. The excess water was primarily used for hygiene-related purposes.However, the difference in time points to the influence of only gross differences in servicelevel, in this case between effectively no access and a service level that can be described asbasic access. Reviewing several studies on water use and collection behaviour, Cairncross(1987) suggests that there is a clearly defined general response of water volumes used byhouseholds to accessibility, shown in figure 2 below.
18
Figure 2.2: Graph of travel time (in minutes) versus consumption (taken from WELL,1998)
Once the time taken to collect water source exceeds a few minutes (typically around 5minutes or 100m from the house), the quantities of water collected decrease significantly.This graph contains a well-defined ‘plateau’ of consumption that appears to operate withinboundaries defined by distances equivalent to around 100 to 1000m or 5 to 30 minutescollection time. There is little change in quantity of water collected within these boundaries(Cairncross and Feachem, 1993). Beyond distance of one kilometre or more than 30 minutestotal collection time, quantities of water will be expected to further decrease, in rural areas toa bare minimum where only consumption needs can be met. In urban areas, where watersupplies may be close but total collection times are very high, greater volumes may becollected that will support hygiene, although the overall impact on household poverty issignificant (Aiga and Umenai, 2002).
Once water is delivered through at least a single tap on-plot, the quantity of water increasessignificantly and further increases are found only when water is piped into the home and isavailable through multiple taps. The findings from a study from Jinja, Uganda, shown in table5 below illustrates this point (WELL, 1998). Average consumption of water when it is pipedinto the home is relatively high (155 l/c/d), but decreases to 50 l/c/d when water is supplied toa yard level. When water is outside the home, average consumption drops still further toroughly one-third the average consumption at a yard tap and one-tenth that of householdswith water piped into the home.
19
Table 5: Average water consumption figures, Jinja, Uganda (WELL, 1998)
Type of supply Averageconsumption (l/c/d)
Service level
Traditional sources,springs or handpumps
15.8 Communal
Standpost 15.5 Communal
Yard tap 50 In compound
House connection 155 Within house(multiple)
The available evidence suggests that the volume of water used in the home is sensitive onlyto gross differences in service level. As noted by WELL (1998) therefore the first priority isto ensure that households reach the plateau (figure 3), that is to have access to an improvedwater source within one kilometre, which corresponds to the current definition of reasonableaccess used in assessing progress in global coverage with water supply and sanitation (WHOand UNICEF, 2000). Beyond this, unless water is provided at a household level, nosignificant changes in water quantities collected will be noted. Clearly benefits at each stageare also accrued from ensuring that the water is of a quality consistent with a tolerable risk tohealth and from ensuring that the water collected is put to effective use for hygiene.
Although there are limited published studies, there appears to be relatively little variation inquantities used when the water is supplied through a yard level of service, probably becausethis level of service does not permit easy use of water-hungry devices and efforts expended toobtain water remain sufficiently high to limit overall quantities used. Once water is suppliedthrough multiple taps within the house, there is more significant variation in volumes used asmore water-hungry and time-saving devices can be deployed and physical effort to obtainwater is largely obviated.
Studies in Kenya, Tanzania and Uganda suggest that the quantities of water used for bathing(including hand washing) and washing of clothes and dishes is sensitive to service level(Thompson et al., 2001). For houses using water sources outside the home, an average of 6.6litres per capita are used for washing dishes and clothes and 7.3 litres per capita for bathing.By contrast for houses with a household connection to piped water supply use on average16.3 litres per capita for washing dishes and clothes and 17.4 litres per capita for bathing. Theauthors suggest that for the households using a water source outside the home, the lesservolume collected has a negative impact on hygiene although this is not quantified.
4.7 Quantity and cost: what influence does this have on use?There have also been suggestions that where water is purchased, the cost may also be alimiting factor on the volumes of water used. The literature is, however, more limited thanthat dealing with the influence of distance. There are suggestions that in some environments agraph of similar dimensions and shape as the quantity-distance graph can be developed(Cairncross and Feachem, 1993; WELL, 1998).
Cairncross and Kinnear (1992) showed that the cost of water purchased from vendors inKhartoum, Sudan, did not led to a significant reduction in the quantity of water procured.Cairncross and Kinnear (1992) suggested that in poorer communities, where an increasing
20
proportion of household income must be spent on acquiring water, the only major item ofexpenditure available for sacrifice was the food budget and therefore it was probable thathigh costs of water were contributing to under-nutrition.
Thompson et al. (2001) note that in East Africa average costs of water for households with apiped water connection in urban areas actually decreased over the preceding 30 years by 20%(when compared to the study by White et al. (1972)). Thompson et al. (2001) also note thataverage costs of water used by households without a household connection in rural and urbanareas increased by 14% and 28% respectively. At the same time, households withconnections to piped water decreased water consumption by 50%, while urban and ruralhouseholds using off-plot water sources actually increased consumption by 60% and 80%respectively.
The overall impact of changes in cost on volumes of water collected does not seem to havebeen significant and may be confounded by other factors. The authors note that the reductionin use by households with a connection to a piped supply may be as a result of a number offactors, including increasing intermittence of piped supplies. A further issue not discussedthat may have influenced quantities used is the expansion in the use of meters leading to morecareful consumption by households. For households lacking a household connection, costincreases may have been off-set by increasing household incomes, allowing greater quantitiesof water to be purchased.
In terms of the influence of cost at a household level, Thompson et al. (2001) indicate thatconsumption in households using off-plot water supplies was strongly influenced byeconomic factors, with wealth of the household being the most important factor, followed bythe cost per litre of water. This showed a marked change from the first study of water usebehaviour in East Africa where container size and source location were the most importantfactors. For households with a connection to the piped water supply, consumption was againnoted as being primarily determined by wealth indicators. In other settings, the increasing useof metering in piped water supplies may influence consumption and the use of progressivetariffs can be applied to penalise excessive use, whilst promoting provision of quantities tomeet basic hygiene needs.
In studies on water usage undertaken in three urban areas in Uganda (Howard et al., 2002)there was limited evidence of a significant association between cost of water and quantities ofwater collected. In one town, Soroti, quantities of water were actually greater from sourceswhere payment was required, although as these were taps supplied to a compound, this islikely to have been a function of access (Howard, 2002). The major influence of the need topurchase water appears to have been to promote multiple source use, thus elasticity was seenprimarily in source selection behaviour rather than reducing volumes or changes in prices.
4.8 Other factors that may affect quantities of water usedFactors such as supply reliability may also influence quantities of water collected, althoughagain there is very limited published data to establish what relationships exist. Zerah (2000)indicates that low-income families in New Delhi are likely to be at greatest risk from poorwater supply continuity. As they have more limited resources, they are less able to store largevolumes of water at home and this led to the use of smaller volumes of water and impairedhygiene, although this is not quantified. It is likely that the nature of the discontinuity willaffect the hardship caused. Whilst regular discontinuity may cause more hardship, this maybe mitigated to some extent if the interruption in supply is predictable as this will allow the
21
household to develop coping strategies for water collection. The greatest problems may befelt when discontinuity is frequent, but very unpredictable. Anecdotal evidence from manyAfrican cities indicate that this is common and may lead to collection of water from pipednetworks at odd hours, including late at night.
When investigating consumption patterns within a piped water supply with multiple levels ofservice, the interaction between the volumes used between different levels of service requiresattention. Volumes of water available for users of lower service levels may be reduced as adirect consequence of over-consumption by households with higher-service levels. This hasbeen shown in a study in Ghana, where low-income communities that relied on public tapsreceived less water and faced greater shortages than high-income communities in partbecause of the consumption patterns of the latter (Stephens, 1996).
A further problem with intermittent water supply is that households may be forced to storewater within or close to the home, thus leading to increasing risks from vector-borne diseasessuch as dengue fever (Ahmad, 2000; Ault, 1994; Rosenbaum et al., 1995). Similar problemswill also noted in houses that only have access to a basic level of service, where water storageis required.
4.9 Laundry: on and off-plot useMinimum requirements for domestic supply should include adequate water for laundry andbathing. In some cases this will be done at the house and in other circumstances some or allof these activities may be carried out at the water source rather than at the household. In bothcases, it would be expected that if an improved source is used, this should provide adequatequantities of water to meet these demands.
Where the source is an improved communal source whether laundering occurs at the house orsource may vary on the nature of the settlement, with greater use at the household expected inurban areas. For instance, a significant proportion of households in urban Uganda appeared tocollect and carry water back to the home for laundering (Howard et al., 2002). In rural areas,it may be socially acceptable for people to bathe and launder clothes at or close to the watersource. In some cases, designs for water supplies include facilities for bathing and laundry.
In some communities, users of communal supplies may be reluctant to transport sufficientwater for laundry and may opt for use of alternative sources. In some communities it has beennoted that the use of water from different source types may vary depending on judgementsmade as to the acceptability of the source for a given use (Madanat and Humplick, 1993). Theterm ‘rationality factor’ has been used to describe this differential use of water from differentsources in rural areas (Almedom and Odhiambo, 1994). For instance, in East Africa it wasnoted that 30% of the population without household connections to a piped water supply useunprotected water sources for laundry (Thompson et al., 2001). This may increase risks tohealth, for instance by increasing exposure to water-based and vector-borne diseases such asschistosomiasis and potentially other vector-borne diseases.
Where water is scarce or beyond the threshold of 1000m, there may be water source or timeconstraints that limit the potential for laundering and bathing at source or transporting waterhome. As a result, the frequency of bathing and laundering may reduce, thus potentiallyincreasing the risks of some infectious diseases (Thompson et al., 2001).
22
4.10 Minimum requirements for all hygiene needsThe evidence suggests that water quantities used by households are primarily dependent onaccess as determined by distance and/or time for collection. These differences are primarilyseen as functioning at four levels, broadly equivalent to service level, shown in table 6 below.The estimated quantities of water at each level may reduce where water supplies areintermittent and the risks of ingress of contaminated water into domestic water supplies willincrease.
Table 6: Service level descriptors of water in relation to hygiene
Service leveldescription
Distance/timemeasure
Likely quantitiescollected
Level of health concern
No access More than 1000mor 30 minutes totalcollection time.
Very low (often lessthan 5 l/c/d).
Very high as hygiene notassured and consumption needsmay be at risk. Quality difficultto assure; emphasis on effectiveuse and water handlinghygiene.
Basic access Between 100 and1000m (5 to 30minutes totalcollection time).
Low. Average isunlikely to exceed20 l/c/d; laundryand/or bathing mayoccur at watersource withadditional volumesof water.
Medium. Not all requirementsmay be met. Quality difficult toassure.
Intermediateaccess
On-plot, (e.g.single tap in houseor yard).
Medium, likely to bearound 50 l/c/d,higher volumesunlikely asenergy/timerequirements stillsignificant.
Low. Most basic hygiene andconsumption needs met.Bathing and laundry possibleon-site, which may increasefrequency of laundering. Issuesof effective use still important.Quality more readily assured.
Optimalaccess
Water is piped intothe home throughmultiple taps.
Varies significantlybut likely above 100l/c/d and may be upto 300l/c/d.
Very low. All uses can be met,quality readily assured..
These levels of access can also be interpreted in terms of household water security. The noaccess group effectively have no household water security as the quantities collected are low,the effort taken to acquire water is excessive and quality cannot be assured. The group withbasic access could be said to also have basic household water security, provided that water is(reasonably) continuous and quality can be assured at source and protected during subsequenthandling.
The group with intermediate access can be said to have effective household water security assufficient water is available to meet domestic needs and quality can be assured. This may beinfluenced by the degree of discontinuity within the supply, but it would be expected thathousehold coping strategies would include bulk storage, which has been shown to beprotective of health in Pakistan (van der Hoek et al., 2002). The group with optimal access
23
also have optimal household water security with quantity, quality and continuity all likely tobe adequate for domestic water needs.
The available data indicates that service level is more relevant to health than quantity ofwater and provides an indication of the volume of water that is available and used byhouseholds. The level of basic access is broadly equivalent to current definitions of minimumquantities are water required for health (WELL, 1998) and intermediate access equivalent toquantity of water that Gleick (1996) argued was the basic water requirement. In relation tocost, the evidence available remains limited and contradictory, as does the evidence forrelationships between reliability and quantity of water. However, it is likely that both willinfluence quantities of water collected in some settings and further research is required in thisarea.
Available evidence suggests that health is significantly compromised at service levelsdescribed as ‘no access’ in table 6 where volumes collected may barely exceed the minimumfor hydration. Estimates suggest that of the global population estimated in 2000 of 6 billionpeople, around 1.1 billion are in a situation categorised as ‘no access’ within this paper.Whilst significant health gains accrue to those 4.9 billion benefiting from ‘basic access’significant health gains continue across categories ‘intermediate access’ and ‘optimal access’.Approximately 2.8 billion of the global population presently have a household connection toa water supply, which covers both the category of intermediate and optimal access.
A minimum for basic health protection corresponds to ‘basic access’ and experience showsthat this is equivalent to a water collection of less than 20 l/c/d, of which about 7.5 litres isrequired for consumption. The effective use in hygiene practices of the limited wateravailable at basic access service level is important if available health benefits are to accrue.The basic level of supply should be regarded as a minimum quantity of water and attentionpaid to increasing levels of service to yard level in order to increase volumes of watercollected.
5 Other Uses of water and links to quantity
5.1 Household productive uses of domestic water
This section deals with the productive uses of domestic water at a household level, whichincludes brewing, small-scale food production and household construction in low-incomeareas. We do not consider community-level enterprises that use water resources in income-generating activities, such as irrigation systems (beyond simple use of water by a householdfor gardening), industry, larger commercial entities, energy production and transport.However, as noted in the introduction, in terms of overall use of water sources the economicuse of water typically greatly exceeds that used for domestic supply, but may compromise theability of the resource to meet basic needs (either through over-consumption or through usesleading to quality deterioration).
The health sector oversight of water supply, has traditionally not considered productive usesof water as important to control. However, it is increasingly recognised that productive usesof water have particular value for low-income households and communities and have healthand well-being benefits (Thompson et al., 2001). Direct health benefits are derived forexample from improved nutrition and food security from gardens crops that have beenwatered. Indirect health benefits arise from improvements in household wealth from
24
productive activity. In urban areas, this often is essential for low-income communities tomeet nutritional requirements and may offer additional income from small-scale sales.
Fass (1993) notes that in families living in 'ultra-poverty' water could form anywhere between1.5 and 10% of the total production costs in household enterprises. The removal of a watersupply, or deterioration in the quality of service, through decreased quantity or availability orincreased intermittence or cost may lead to further poverty among poor households using thiswater for small-scale economic activities such as food production. The quality of water usedfor productive processes needs to be suitable for domestic supply where it is used forprocessing food for retail or in some circumstances irrigation for its production.
5.2 Amenity uses of water
Amenity uses are not typically considered in relation to health aspects of water quantity.Amenity uses include lawn-watering and car washing, although in some cases the latterwould be more correctly categorised as productive uses of water as it may be used to providean income. There are some benefits of purely amenity uses of water in terms of quality oflife. However, particularly for the most vulnerable, amenity use of water is likely to belimited.
The principal concern in relation to amenity uses of domestic water supplies is to reduce theconsumption of water for these purposes when this may place a significant demand on thewater supply such that universal basic access is compromised. This is a problem noted inmany developed countries where increasing efforts are being made to educate users on theproblems of use of water for such purposes and in some cases restrictions applied andenforced to conserve water resources. Such approaches may include the use of variable tariffrates that penalise excess use, although in this case care must be taken to ensure that therewould be no financial penalty for ensuring water for basic needs, including laundry.
Over-use of scarce water supplies for amenity uses is also found in developing countries,particularly in urban areas where the patterns of use among the wealthy may directly impacton the availability of water to the poor (Stephens, 1996). Therefore controlling amenity use ofdomestic water supplies should be driven to ensure that basic needs are met throughout thepopulation in an equitable manner. It is important that such controls consider not onlymeeting the basic needs for current populations, but also takes into account future populationgrowth. Furthermore, as many developing countries are either already experiencing or facingwater scarcity and water stress, the need to control consumption of water to conserveresources is also critical (Gleick, 1993).
6 ImplicationsThe evidence reviewed and conclusions drawn have significant implications regarding thedevelopment and application of norms for water quantity. This includes advocacy for accessnot quantity, the monitoring and evaluation with progress in meeting international andnational goals for water supply, prioritising interventions and applications within specificcontexts, such as emergencies.
6.1 Changing the nature of the debate: household water security/access not quantityThe available evidence indicates that the quantity of water that households collect and use isprimarily dependent on accessibility (as determined by both distance and time). There is
25
some indication that cost and reliability may also influence quantity of water collected,although the available evidence is limited and often contradictory.
The influence of accessibility operates as a function of service level, which can be dividedinto the five categories shown table 6 above. It is therefore logical that the debate regardingquantity is not related to volumes of water available but by the level of service provided.Increases in quantities of water used will only be achieved through upgrading of servicelevel. Furthermore household water security improves with increasing service level, whichwill contribute to reducing poverty.
The first priority for interventions to improve access to water supplies is to ensure that at leastbasic access is achieved. At a basic level of service the volume of water collected is likely tobe around 20 litres per capita per day. There is no evidence that behaviour changeinterventions have been successful in promoting increases in water quantities used inhouseholds that have achieved basic access to improved water supply. At this level of service,it is the effective use of the available water that is of principal importance, including theimportance and timing of hand and face-washing and household water treatment, incontrolling infectious disease transmission. Once this level of access is achieved, attentionshould be placed on providing guidance and support in hygiene practices and water qualitymanagement techniques that will reduce the risk of diarrhoeal disease transmission.
Notwithstanding the emphasis on personal hygiene, further health gains are typicallyassociated with increasing levels of service - especially in upgrading to an ‘on plot’ level ofaccess and to a lesser extent when upgrading from this level to ‘multiple tap in house’.Provision of ‘intermediate access’ levels of service will result in the use of approximately 50litres of water per capita per day. ‘Optimal access’ will result in much higher quantities ofwater being consumed and the emphasis may change to restriction on quantity used.Upgrading to these levels of service therefore represent successive priorities, once basicaccess is achieved.
Overall, many of the health benefits ultimately accrue from proper water usage and goodhygiene behaviours and simple provision of infrastructure alone is unlikely to maximisehealth gains. Poor hand-washing practice has been reported in countries with abundant in-house water supplies and general availability of soap and interventions to improve hygienewill usually provide some improvements in health in any setting, irrespective of socio-economic conditions or service level.
In many developing countries, the challenge remains to improve overall system managementto reduce intermittence, make costs of water affordable and to develop incentives for users tobe willing to obtain a legal connection and to (Howard, 2001; Sansom et al., 2000; WELL,1998; World Bank, 1993). It may also require rethinking how services can be delivered to theurban poor in particular and developing approaches to offering a range of technical options tocommunities, community management and more effective financing of sector improvements(Briscoe, 1996; Cotton and Tayler, 1994; Howard, 2001; Subramanian et al., 1997; WorldBank, 1993).
6.2 International Development TargetsThe UN (2000) set a Millennium Development Goal to 'halve the proportion of people whoare unable to reach or to afford safe drinking water by 2015'.
26
Ensuring access to the basic level of service represents the primary objective of theMillennium Development Goal in relation to water, although the definition of the safety ofwater remains unclear. In 2000, this target was realised for the 82% of the global populationwho had a improved water source within one kilometre of their home or a householdconnection to a water supply. The remaining 18% (equivalent to about 1.1 billion people) arein the category of no access and are found largely in Asia and Africa and particularly in ruralpopulations which are typically less well served than urban populations (WHO and UNICEF,2000). This suggests a priority on ensuring at least access basic access in rural areas. It isimportant to note, however, that the rapid growth in urban populations suggests that figuresfor access in urban areas may reduce over time unless the pace of expansion of coverage ismaintained (WHO and UNICEF, 2000).
Where at least basic access is already assured for all, a priority is to increase the numbers ofhouseholds that reach the intermediate access level, where the evidence suggests furthermajor improvements in health are achieved (Prüss et al., 2002). Currently, only 47% of theglobal population has access to this level of service, a figure significantly lower than globalaccess to improved sanitation of 60% (WHO and UNICEF, 2000).
Increasing numbers of households in this category will result in significant public healthgains from likely better hygiene, maintenance of microbial quality of water and more time forchild-care, productive activity and schooling. Such access is likely to also require that thesupply is continuous, although there remains a dearth of direct evidence of the overall impactof intermittence on public health (Prüss et al., 2002), although the risks in relation tocontamination of supplies are well known (Regli et al., 1993; Clark et al., 1993). In manyurban situations, it may be possible to move from no access to intermediate access without anintermediate phase of basic service level. Responding to local conditions and opportunitieswill be more effective than developing generic blue-prints for service level upgrading.
The achievement of an intermediate level of water supply service may create additionalproblems that will need to be addressed to maximise public health gains. Most notable are thedisposal of sullage and improvements in drainage around individual dwellings and incommunities. The presence of poorly drained water may lead to increased risks from insectvector-related diseases, may increase the risk of exposure to pathogens, particularly for smallchildren who may play near or in water. It could potentially increase risks of schistosomiasis,which has been noted as an increasing problem in poorly drained urban areas (WHO, 1991).
Ensuring basic access to the currently unserved and increasing the numbers of people withintermediate access are complementary activities. There remains no doubt that ensuring atleast a basic level of service remains a key international goal. At the same time, investmentshould not focus solely on this level of access, but should also be targeted on addressingmoving increasing numbers of people to an intermediate level of access.
6.3 Global AssessmentIn the Global Water Supply and Sanitation Assessment 2000 Report, improved sanitationtypes were all defined primarily at individual household levels, or at most shared by a fewneighbouring households. By contrast, most definitions of improved water supply weredefined as communal (basic access) levels of service.
If the definition of improved water supply were on-plot (intermediate access), which can bejustified on the basis on health evidence, water supply access is actually lower than sanitation
27
access. This has profound implications. Currently, there is rightly significant advocacy toreduce the sanitation access deficit. However, the evidence suggests that continued attentionis required equally for improving services in both water supply and sanitation.
6.4 Emergencies and disastersIn emergency/disaster situations the provision of adequate safe water is of particular concernin order to control the spread of infectious disease and in particular to prevent large-scaleoutbreaks of disease. The provision of sufficient safe water for consumption and cooking isan absolute requirement and should never be compromised.
The reference volume of 7.5 litres for consumption must be available for use within thedwellings within an emergency context and the may the first goal in the first stages of anemergency response. Short term decreases in volumes available for hygiene purposes mayhave limited health implications provided effective use in hygiene, particularly handwashing,is promoted.
The emergency response should include design to provide water for hygiene and it would berational to consider this in terms of service level in the same way as non-emergencysituations. In terms of estimating source volume requirements, it can be assumed thatprovision of communal facilities will result in use of on average 20 litres per capita per day, iftaps are provided to individual dwellings this will increase to 50 litres per capita per day andif multiple taps are provided this will increase further.
6.5 Quantity in health-based surveillance programmesThe role of the health sector in promoting better water supplies to maximise health gainsshould be to advocate for those interventions that will deliver greatest improvements inhealth. A key component of this role is the regular surveillance of water supplies in order toassess progress with meeting targets consistent with stated public health goals and to identifypriority areas for intervention. WHO (1993) define surveillance as ‘the continuous andvigilant oversight of drinking water supplies from a public health perspective’.
Water quantity has traditionally be viewed as forming a key indicator for such surveillanceprogrammes, alongside measures of quality, cost, continuity and coverage (Howard, 2002 ;Lloyd and Bartram, 1991; Lloyd and Helmer, 1991; WHO, 1997). Elsewhere it has beensuggested that service level may be more appropriate (Bartram, 1999).
This review indicates that service level is more closely related to health than the quantity ofwater and also provides an indication of the likely volumes of water available and used byhouseholds. Quantity data is also difficult to collect and are often aggregated (for instance atcommunity level) whereas water quantity or availability is experienced at the householdlevel (Bartram 1999). There may be a value in assessing volumes used by households withhigher service levels, in part to determine the influence their consumption has on users atbasic level services. For such an approach to be useful, however, this must be matchedagainst a set of surrogate variables – for instance distance, time, cost or reliability – in orderfor conclusions about a wider population to be drawn.
The literature suggests that some evaluation of the impact of cost and reliability of watersupply on quantities of water used would have value within a surveillance programme,although this would not be expected to require frequent monitoring (Howard, 2002; Lloyd
28
and Bartram 1991). There appears little justification for collection of such data in relation todistance, as the literature points to this only being significant between gross differences inservice level and of limited value for assessing within-group variation at basic andintermediate level of service.
This indicates that estimates of access as described by service level and continuity in supplyprovide the most meaningful insights into the adequacy of water supply for health protectionand can provide an indication of the quantities of water likely to be collected and used. Thisdata is easily obtainable using inventories of water sources and connections, sanitaryinspection of water supplies and water usage studies. There may be some value in collectingdata on quantities of water collected in relation to cost, although the likelihood of relationshipmay depend on the availability of different types of water source. In situations where vendorusage is more common there would appear to be a justification of assessments of volumes ofwater collected.
7 AcknowledgementsThe authors would like to thank the following for their constructive comments during thepreparation of this document:
Andy Bastable (OXFAM); Hamid Bakir (WHO-CEHA); Sandy Cairncross (London Schoolof Hygiene and Tropical Medicine); Susan Murcott (Massachusetts Institute of Technology);Amaka Obika (WEDC).
WHO Geneva: Richard Carr, Jose Hueb, Roisin Rooney.
8 ReferencesAiga H and Umenai T, 2002, Impact of improvement of water supply on household economyin a squatter area of Maila, Social Science and Medicine, 55(4): 627-641.Al-Ali F M, Hossain M M and Pugh R N H, 1997, The associations between feeding modesand diarrhoea among urban children in a newly developed country, Public Health, 111: 239-243.Almedom A and Odhiambo C, 1994, The rationality factor: choosing water sources accordingto uses, Waterlines, 13(2): 28-31.Bailey R, Downes B, Downes R and Mabey D, 1991, Trachoma and water use; a case-controlstudy in a Gambian Village, Transactions of the Royal Society of Tropical Medicine andHygiene, 85: 824-828.Baltazar J C, Tigalo T V and Tempongko S B, 1993, Hygiene behaviour and hospitalisedsevere childhood diarrhoea: a case-control study, Bulletin of the World Health Organization,71(3/4): 323-328.Bartram J, 1999. Effective Monitoring of Small Drinking-water Supplies in Providing SafeDrinking-water in Small Systems, Cotruvo A, Craun G and Hearne N (Eds). CRC Press, BocaRaton, USA, pp: 353-365.Birmingham M E, Lee L A, Ntakibirora M, Bizimana F and Deming M S, 1997, A householdsurvey of dysentery in Burundi: implications for the current pandemic in sub-Saharan Africa,Bulletin of the World Health Organization, 75(1): 45-53.Bradley D, 1977, Health aspects of water supplies in tropical countries, In Feachem R,McGarry M and Mara D, Water, wastes and health in hot climates, John Wiley and Sons,Chichester, UK, pp 3-17.
29
Briscoe J, 1996, Financing water and sanitation services: the old and new challenges, WaterSupply, 14(3/4): 1-17.Bukenya G B and Nwokolo N, 1991, Compound hygiene, presence of standpipe and risk ofchildhood diarrhoea in an urban settlement in Papua New Guinea, International Journal ofEpidemiology, 20(2): 534-539.Buttle M and Smith M,1996, Out in the cold: emergency water supply and sanitation in coldclimates. WEDC, Loughborough University.Cairncross S, 1987, The benefits of water supply, In Pickford, J (ed) Developing WorldWater. Grosvenor Press, London.Cairncross A M, 1990, Health impacts in developing countries: new evidence and newprospects, Journal of the Institution of Water and Environmental Management, 4(6): 571-577.Cairncross S, 1993, Control of enteric pathogens in developing countries, In Mitchell R (ed)Environmental microbiology. Wiley-Liss, New York, pp157-189.Cairncross S and Kinnear J, 1992, Elasticity of demand for water in Khartoum, Sudan, SocialScience and Medicine, 34(2): 183-189.Cairncross S and Feachem R, 1993, Environmental health engineering in the tropics: anintroductory text (2nd edition). John Wiley and Sons, Chichester, UK.Carter R C, Tyrrel S F and Howsam P, 1997, The impact and sustainability of water andsanitation programmes in developing countries, Journal of the Chartered Institution of Waterand Environmental Management, 13: 292-296.Chan J, Knutsen S F, Blix G G, Lee J W and Fraser G E, 2002, Water, other fluids, and fatalcoronary heart disease, American Journal of Epidemiology, 155(9): 827-833.Clark R M, Goodrich J A and Wymer L J, 1993, Effect of the distribution system ondrinking-water quality, Journal of Water Supply Research and Technology – Aqua, 42(1): 30-38.Cotton A P and Tayler W K, 1994, Community management of urban infrastructure indeveloping-countries, Proceedings of the Institution of Civil Engineers-Municipal Engineer,103(4): 215-224.Curtis V, Cairncross S and Yonli R, 2000, Domestic hygiene and diarrhoea - pinpointing theproblem, Tropical Medicine and International Health, 5(1): 22-32.Dufault A, 1988, Women carrying water: how it affects their health, Waterlines, 6(3): 23-25.Emerson P M, Cairncross S, Bailey R L and Mabey D C W, 2000, Review of the evidencebase for the ‘F’ and ‘E’ components of the SAFE strategy for trachoma control, TropicalMedicine and International Health, 5(8): 515-527.Esrey SA, Feachem RG and Hughes JM, 1985, Interventions for the control of diarrhoealdiseases among young children: improving water supplies and excreta disposal facilities,Bulletin of the World Health Organization, 63(4): 757-772.Esrey SA, Potash JB, Roberts L and Shiff C, 1991, Effects of improved water supply andsanitation on ascariasis, diarrhoea, dracunculiasis, hookworm infection, schistosomiasis, andtrachoma, Bulletin of the World Health Organization, 69(5): 609-621.Esrey SA, Habicht J-P and Casella G, 1992, The complementary effects of latrines andincreased water usage on the growth of infants in rural Lesotho, American Journal ofEpidemiology, 135(6): 659-666.Esrey S, 1996, Water, waste and well-being: a multi-country study, American Journal ofEpidemiology, 143(6): 608-623.Fass S M, 1993, Water and poverty: implications for water planning, Water ResourcesResearch, 29(7): 1975-1981.Food and Nutrition Board, 1989, Recommended dietary allowances 10th edition, NationalAcademy Press, Washington, DC.
30
Fewtrell L and Bartram J (Eds) Water Quality: Guidelines, Standards and Health, WorldHealth Organization, IWA Publishing, London, UK.Ford T E, 1999, Microbiological safety of drinking water: United States and globalperspectives, Environmental Health Perspectives, 107 (S1): 191-206.Ghosh S, Sengupta P G, Mondal S K, Banu M K, Gupta D N and Sircar B K, 1997, Riskbehavioural practices of rural mothers as determinant of childhood dairrhoea, Journal ofCommunicable Diseases, 1: 7-14.Gibbs S, 1996 Skin disease and socio-economic conditions in rural Africa: TanzaniaInternational Journal of Dermatology 35(9): 633-639.Gilman R H, Marquis G S, Ventura G, Campos M, Spira W and Diaz F, 1993, Water cost andavailability: key determinants of family hygiene in a Peruvian shantytown, American Journalof Public Health, 839(1): 1554-1558.Gleick P H (ed.), 1993, Water in crisis: a guide to the World's freshwater resources. OxfordUniversity Press, New York, USA.Gleick P H, 1996, Basic water requirements for human activities: meeting basic needs, WaterInternational, 21: 83-92.Gorter A C, Sandiford P, Smith G D and Pauw J P, 1991, Water supply, sanitation anddiarrhoeal disease in Nicaragua: results from a case-control study, International Journal ofEpidemiology, 20(2): 527-533.Grandjean A C, Reimers K J, Bannick K E and Haven M C, The effect of caffeinated, non-caffeinated, caloric and non-caloric beverages on hydration, Journal of the American Collegeof Nutrition, 19(5): 591-600.Haas C N, Rose J B and Gerba C P, 1999, Quantitative microbial risk assessment, Wiley,New York.Herbert J, 1985, Effects of components of sanitation on nutritional status: findings fromSouth Indian settlements, International Journal of Epidemiology, 14(1): 143-151.Hoque B A and Briend A, 1991, A comparison of local handwashing agents in Bangladesh,Journal of Tropical Medicine and Hygiene, 94: 61-64.Hoque B A, Mahalanabis D, Alam M J and Islam M S, 1995, Post-defecation handwashing inBangladesh: practice and efficiency perspectives, Public Health, 109: 15-24Howard G, 2001, Challenges in increasing access to safe water in urban Uganda: economic,social and technical issues, In Craun, G F, Hauchman F S and Robinson DE, Safety of WaterDisinfection: balancing microbial and chemical risks, ILSI Publications, Washington, USA,pp 483-499.Howard G, 2002, Effective Approaches to Water Supply Surveillance in Urban Areas ofDeveloping Countries Guildford, Surrey, Surrey University, PhD Thesis.Howard, G (ed), 2002, Water supply surveillance - a reference manual. WEDC,Loughborough University, UK.Howard G, Teuton J, Luyima P and Odongo R, 2002, Water usage patterns in low-income urbancommunities in Uganda: implications for surveillance, International Journal of EnvironmentalHealth Research, 12(1): 63-73.Howard G and Pond K, 2002, Drinking water surveillance programmes in the South-EastAsia Region: updated situation assessment and recommendations for future activity.WHO/SEARO, New Delhi, India.Hsieh Y-H, Bobo L D, Quinn T C and West S K, 2000, Risk factors for trachoma: 6-yearfollow-up of children aged 1 and 2 years, American Journal of Epidemiology, 152(3): 204-211.Huttly S R A, Morris S S and Pisani V, 1997, Prevention of diarrhoea in young children indeveloping countries, Bulletin of the World Health Organization, 75(2): 163-174
31
IPCS, 1994. Environmental Health Criteria 170:Assessing human health risks ofchemicals:derivation of guidance values for health-based exposure limites. WHO Geneva.Jackonen S, 1997, Dehydration and hydration in the terminally ill: care considerations,Nursing Forum, 32(3): 5-13.Khan MU, 1982, Interruption of shigellosis by handwashing, Transactions of the RoyalSociety of Tropical Medicine and Hygiene, 76(2): 164-8Kleiner S M, 1999, Water: an essential but overlooked nutrient. Journal of the AmericanDietetic Association, 99 (2); 200-206.Lloyd B and Helmer R, 1991, Surveillance of drinking water quality in rural areas. Longman,London.Lloyd B and Bartram J, 1991. Surveillance Solutions to Microbiological Problems in WaterQuality Control in Developing Countries. Water Sci Tech 24(2): 61 – 75.Luna E J A, Medina N H, Oliveira M B, De Barros O M, Vranjac A, Melles H H B, West Sand Taylor H R, 1992. Epidemiology of trachoma in Bebeduro State of Sao Paulo, Brazil:Prevalence and risk factors, International Journal of Epidemiology, 21(1): 169-177.Mackenzie W R, Hoxie N J, Proctor M E, Gradus M S, Blair K A, Peterson D E,Kazmierczak J J, Addiss D G, Fox K R, Rose J B and Davis J P, 1994, A massive outbreak inMilwaukee of Cryptosporidium infection transmitted through the public water supply, NewEngland Journal of Medicine, 331: 161-167.Madanat S and Humplick F, 1993, A model of household choice of water supply systems indeveloping countries, Water Resources Research, 29(5): 1353-1358.Mara D D and Cairncross S, 1989, Guidelines for the safe use of wastewater and excreta inagriculture and aquaculture: measures for public health protection, WHO, Geneva,Switzerland.Melian R, Mrlian N, Gouriev A, Moraru C and Radstake F, Groundwater quality and ruraldrinking-water supplies in the Republic of Moldova, Hydrogeology Journal, 7:188-196.National Institutes of Health, 2002, Medical Encyclopedia,http://www.nlm.nih.gov/ medlineplusNeave N, Scholey A B, Emmett J R, Moss M, Kennedy D O and Wesnes K A, 2002, Wateringestion improves subjective alertness, but has no effect on cognitive performance indehydrated healthy young volunteers, Appetite, 37(3): 255-256.O'Connor R, 2002, Report of the Walkerton Enquiry, Part One: the events of May 2000 andrelated issues (a summary), Ontario Ministry of the Attorney General, Ontario, Canada.Oo K N, Aung W W, Thida M, Toe M M, Lwin H H and Khin E E, 2000, Relationship ofbreast-feeding and hand-washing with dehydration in infants with diarrhoea due toEscherichia coli. Journal of Health, Population and Nutrition, 18(2): 93-96.Page B, 1996, Taking the strain - the ergonomics of water carrying, Waterlines, 14(3): 29-31.Payment P and Hunter P R, 2001, Endemic and epidemic infectious intestinal disease and itsrelationship to drinking water, In Fewtrell L and Bartram J (eds.) Water quality: guidelines,standards and health, IWA, London, pp 61-88.Petersen E A, Roberts L, Toole M J and Peterson D E, 1998, The effect of soap distributionon diarrhoea: Nyamithuthu refugee camp, International Journal of Epidemiology, 27: 520-524.Phillips P A, Rolls B J, Ledingham J G, Forsling M L, Moton J J, Crowe M J et al.et al..,1984, Reduced thirst after water deprivation in healthy elderly men, New England Journal ofMedicine, 311: 753-759.Pinfold J, 1990, Faecal contamination of water and fingertip rinses as a method for evaluatingthe effect of low-cost water supply and sanitation activities on faeco-oral diseasetransmission. II. A hygiene intervention study in rural north-east Thailand, Epidemiology andInfection, 105: 377-389.
32
Prost A and Négrel A D, 1989, Water, trachoma and conjunctivitis, Bulletin of the WorldHealth Organization, 67(1): 9-18.Prüss A and Mariotti S P, 2000, Preventing trachoma through environmental sanitation: areview of the evidence base, Bulletin of the World Health Organization, 78(2): 258-264.Prüss A and Havelaar A, 2001, The global burden of disease study and applications in water,santiation and hygiene, In Fewtrell L and Bartram J (eds) Water quality: guidelines,standards and health, IWA, London, pp 43-59.Prüss A, Kay D, Fewtrell L and Bartram J, 2002, Estimating the burden of disease fromwater, sanitation and hygiene at a global level, Environmental Health Perspectives, 110(5):537-542.Regli S, Rose J B, Haas C N, Gerba C P 1991 Modeling the risk from giardia and viruses indrinking water Journal of the American Water Works Association 83(11): 76-84.Rogers P J, Kainth A and Smit H J, 2001, A drink of water can improve or impair mentalperformance depending on small differences in thirst, Appetite, 36(1): 57-58.Rowe J W, Shock N W and DeFronzo R A, 1976, The influence of age on the renal responseto water deprivation in man, Nephron, 17: 270-278.Roy R N and Sinclair J C, 1975, Hydration of the low birth-weight infant, Clinics inPerinatology, 2(2): 393-417.Sansom K, Coates S, Njiru C, Franceys R, 2000 Strategic marketing to improve both waterutility finances and services to poor urban consumers WEDC Discussion Paper WEDC,Loughborough.Saywell D, 1999, Pollution from on-site sanitation - the risks? What risks? Waterlines, 17(4):22-23.Sazawal S, Bhan M J K, Clemens J and Bhatnagar S, 1991, Evidence for recent diarrhoealmorbidity as a risk factor for persistent diarrhoea: a case control study, International Journalof Epidemiology, 20(2): 540-545.Shahid N S, Grenough W B, Samadi A R, Huq M I and Rhamn N, 1996, Hand washing withsoap reduces diarrhoea and spread of bacterial pathogens in a Bangladesh village, Journal ofDiarrhoeal Disease Research, 14(2): 85-89.Sircar B K, Sengupta P G, Modal S K, Gupta D N, Saha N C, Ghosh S, Deb B C and Pal S C,1987, Effect of handwashing on the incidence of diarrhoea in a Calcutta slum, Journal ofDiarrhoeal Disease Research, 5(2): 112-114.Smith A H, Lingas E O and Rahman M, 2000, Contamination of drinking-water by arsenic inBangladesh: a public health emergency, Bulletin of the World Health Organization, 78(9):1093-1103.Sobsey M, 2002, Managing Water in the Home: accelerated health gains form improvedwater supply, WHO, Geneva, Switzerland.Sphere Project, 2002, Humanitarian charter and minimum standards in disaster response,OXFAM Publishing, Oxford, UK.Stanton B F, Clemens J D, 1987 An education intervention for altering water-sanitationbehavior to reduce childhood diarrhea in urban Bangladesh II. A randomised trial to assessthe impact of intervention on hygienic behaviors and rates of diarrhea American Journal ofEpidemiology 125(2): 292-301.Stephens C, 1996 Healthy cities or unhealthy islands? The health and social implications ofurban inequality Environment and Urbanization 8(2): 9-30.Subramanian A, Jagganathan N V and Meizen-Dick R, 1997, User Organizations forsustainable water services, World Bank, Washington DC.Thompson J, Porras I T, Tumwine J K, Mujwahuzi M R, Katui-Katua M, Johnstone N andWood L, 2001, Drawers of Water II: 30 years of change in domestic water use andenvironmental health in East Africa, IIED, London, UK.
33
UNDP, 1999, Human Development Report 1999, Oxford University Press, New York, USA.United Nations General Assembly, 2000 Resolution 55/2: Millennium Declaration UnitedNations, New York.United States Army Center for Health Promotion and Preventive Medicine, 2003,http://chppm-www.apgea.army.mil/.VanderSlice J and Briscoe J, 1995, Environmental interventions in developing countries:interactions and their implications, American Journal of Epidemiology, 141(2): 135-141.Van der Hoek W, Feenstar SG and Konradsen F, 2002, Availability of irrigation water fordomestic use in Pakistan: its impact on prevalence of diarrhoea and nutritional status ofchildren, Journal of Health, Population and Nutrition, 20(1): 77-84.WELL 1998, Guidance manual on water supply and sanitation programmes, WEDC,Loughborough, UK.West S, Lynch M, Turner V, Munoz B, Rapoza P, Mmbaga B B O and Taylor H R, 1989,Water availability and trachoma, Bulletin of the World Health Organization, 61(1): 71-75.West S, Munoz B, Turner V M, Mmbaga B B O and Taylor H R, 1991, The epidemiology oftrachoma in central Tanzania, International Journal of Epidemiology, 20(4): 1088-1092.West S, Munoz B, Lynch M, Kayongoya A, Mmbaga B B O and Taylor H R, 1996, Riskfactors for constant, severe trachoma among pre-school children in Kongwa, Tanzania,American Journal of Epidemiology, 143(1): 73-78.White G F, Bradley D J, White A U, 1972, Drawers of water: domestic water use in EastAfrica, University of Chicago Press, Chicago.WHO, 1985 Guidelines for drinking-water quality: Volume 1 Recommendations 1st edition.WHO, Geneva., Switzerland.WHO, 1991, Surface water drainage for low-income communities, WHO, Geneva,Switzerland.WHO, 1993, Guidelines for drinking-water quality: Volume 1 Recommendations 2nd edition,WHO, Geneva (2nd edition) Switzerland.WHO, 1996, Guidelines for Drinking Water Quality, Volume 2: Health and SupportingCriteria 2nd edition, WHO, Geneva, Switzerland.WHO, 1997, Guidelines for drinking-water quality, Volume 3: Surveillance and control ofcommunity water supplies 2nd edition, WHO, Geneva, Switzerland.WHO, 1999, Fluoride in drinking water (draft for consultation), WHO, Geneva, Switzerland.WHO, 2000, Health systems: improving performance, World Health Report, 2000, WHO,Geneva, Switzerland.WHO, in finalisation, Guidelines for safe recreational water environments, WHO, Geneva,Switzerland.WHO and UNICEF, 2000, Global Water Supply and Sanitation Assessment 2000 Report,WHO/UNICEF, Geneva/New York.World Bank, 1993, Water resources management, World Bank, Washington DC, USA.Zerah M-H, 2000, Household strategies for coping with unreliable water supplies: the case ofDehli, Habitat International, 24: 295-307.Zerihun N, 1997, Trachoma in Jimma Zone, South Western Ethiopia, Tropical Medicine andInternational Health, 2(12): 1115-1121.
